ECREA16-Abstract_book

549 / 658

547
Saturday, November 121 1 : 0 0 – 1 2 : 3 0
POL17 Selective Exposure
PP 593 Still an Agenda Setter? The Traditional News Media and Public Opinion from Low-Choice to High-Choice Environments: A Longitudinal
Study
A. Shehata1, M. Djerf-Pierre1
1 University of Gothenburg, Department of Journalism- Media and Communication, Gothenburg, Sweden
In light of the tremendous media environmental transformations that signifies contemporary democracies, the relevance of classic media effects theories
are now frequently questioned (Bennett & Iyengar, 2008; Holbert, Garrett & Gleason, 2010). With an increasing number of media outlets and growing op‑
portunities for media choice, aggregate public opinion should be less responsive to the agenda of the traditional news media, it is argued. Several scholars
have pointed to the fragmentation of the media landscape as well as how citizens increasingly select media content according to personal interests and
preferences – even “avoiding” news about politics and current affairs altogether (Chaffee & Metzger, 2001; Prior, 2007). Thus far, however, few attempts
have been made to confront the argument that the traditional media has gradually lost its impact on public opinion along the path from a low-choice to
a high-choice media environment. In the present study we seek to do so by focusing on one of the media effect theories at the heart of this discussion:
agenda setting theory (Chaffee & Metzger, 2001; Shehata & Strömbäck, 2013). More specifically, this study combines annual nationally representative
public opinion surveys conducted in Sweden with media content analysis, covering a period of almost 30 years, from 1986 to 2014, in order to analyze
whether public perceptions of issue importance have become less sensitive to fluctuations in their salience in the traditional media. By doing so, we achieve
several things. First, we are able to capture the relationship between the media and public agenda over time, making it possible to assess the frequent
argument of a waning impact of the traditional news media. Second, we bring the study of media effects back to a natural setting, thereby complementing
the important – and ever more sophisticated – experimental studies that have come to dominate the field in recent years. We argue that the external
validity of this experimental research in times of abundant media choice, must necessarily be assessed with reference to evidence from“the real world”: To
what extent does aggregate public opinion still respond to shifts in the agenda of traditional news media? Third, the study also seeks to contribute to our
understanding of the contingency of agenda setting effects. In particular, our extensive longitudinal design enables a thorough analysis of how the salience
of different issues in the traditional media influence the public. But rather than focusing on characteristics of the issues themselves only, such as their
obtrusiveness (Soroka, 2002; Zucker, 1978), we also analyze characteristics of the media agenda in order to critically examine the concept of media salience
– which has been far from clear-cut in agenda setting research (Kiousis, 2004). Based on previous research we suggest two dimensions of the media agen‑
da, which we believe influence the transfer of issue salience from the media to the public: (1) issue competition and (2) issue history. In sum, the findings
indicate that the traditional news media can still excert agenda setting influence on public opinion, but these effects have weakened over time.
PP 594 Selective Online Exposure and Political Polarization During Swedish Election Campaigns: A Longitudinal Analysis Using Four Waves
of Panel Data
P.M. Dahlgren1, A. Shehata1, J. Strömbäck1
1 Gothenburg University, Journalism- Media and Communication, Gothenburg, Sweden
The rapidly changing media environment has dramatically altered how citizens can access political news and form opinions. With growing opportunities
for media choice, people can increasingly select media based on already held political ideologies. This is particularly true in an online environment where
the opportunities to find attitude-consistent – and avoid attitude-inconsistent – political news are almost endless (Garrett, 2009). Based on selective expo‑
sure theory, some have argued that citizens will increasingly use online media selectively in a way that strengthens their already held beliefs and polarizes
public opinion (Bennett & Iyengar, 2008). While there are some American studies supporting this claim (e.g., Stroud, 2010), political scientist Markus Prior
recently concluded that the“evidence for a causal link betweenmore partisanmessages and changing attitudes or behaviors is mixed at best”(2013, p. 101).
In addition, research on selective exposure outside the US context is limited, and the evidence for selective exposure to online news outside the US context
is virtually non-existent. How selective exposure and polarization generalizes to other countries is unclear given the significant differences among political
systems and media systems in the US and most European countries (Hallin & Mancini, 2004). Against this background, we analyze ideological selective
exposure to online news and political polarization in Sweden. Based on a representative four-wave panel study (n=2,283) conducted during the 2014
Swedish national election campaign, we study the mutual reinforcement between political ideology and online media use over a period of 6 months.
More specifically, we hypothesize that Internet users primarily exposed to attitude-consistent online news strengthen their preexisting political ideology
over time, leading to political polarization (H1), while exposure to attitude-inconsistent news alleviates political polarization (H2). This study contributes
to the literature on selective exposure and polarization in at least four ways. First, a large-scale national-sample survey helps generalize the findings to
a broader population. Second, multiple exposures over time increase the ability to infer a causal relationship compared to single exposure or cross-sectional
data. Third, testing selective exposure in Sweden’s multi-party system is likely to reveal less dichotomized (and thus less polarized) results compared to
the binary two-party system of the US. Forth and lastly, we investigate Internet, an environment where selective exposure to attitude-consistent news is
particularly relevant due to the excess choice individuals are confronted with. References Bennett, W. L., & Iyengar, S. (2008). A New Era of Minimal Effects?
The Changing Foundations of Political Communication. Journal of Communication, 58(4), 707–731. –2466.2008.00410.x
Garrett, R. K. (2009). Politically Motivated Reinforcement Seeking: Reframing the Selective Exposure Debate. Journal of Communication, 59(4), 676–699.
–2466.2009.01452.x Prior, M. (2013). Media and Political Polarization. Annual Review of Political Science, 16(1), 101–127.
135242 Stroud, N. J. (2010). Polarization and Partisan Selective Exposure. Journal of Communication,
60(3), 556–576. 2466.2010.01497.x